    [Patent document 1] JP-H11-328596 A
In recent years, there is known a vehicular application which provides a driving assistance or control of a behavior in cooperation with an in-vehicle navigation apparatus. Such an application is referred to as a navigation-cooperative application. Such a navigation-cooperative application uses map data or a present position acquisition function, which is provided in the navigation apparatus, and serves as an apparatus or function to provide an assistance to a safe and comfortable driving operation. For example, Patent document 1 discloses a road shape learning apparatus provided as follows. The maximum radius of a curvature of a travel line expected in a curve is acquired based on road shape data such as a road width or curvature radius of a curve out of various data contained in map data. Based on the acquired maximum curvature radius, a target vehicle velocity at the curve entrance is calculated.
In Patent document 1, the road shape of the travel line is acquired from the map data; a suitable vehicle velocity is designated according to the road shape. This enables an enhancement in safety and a comfortable travel meeting with driver's sensibility.
Incidentally, there is a case that map data is prepared by old measurement data or an error occurs when authoring map data. Such a case may pose a risk in the configuration where a target vehicle velocity is calculated based on the curvature radius stored in the map data like in Patent document 1. That is, when the curvature radius greater than that of the actual shape of the road is stored in the map data, there is a possibility that the target vehicle velocity is calculated to be higher than an intended one. Therefore, in such a navigation-cooperative application, especially, to control the behavior of the vehicle, the detailed map data is thereby necessary which indicates the shape of the road accurately.
However, in order to prepare the detailed map data which indicates the shapes of roads accurately, a huge cost and large volume of storage area are needed. It is not easy to previously prepare the map data accurately.